Data and classificaiton processes used

data= pd.read_csv("C:\\data_set\\diabetes.csv")
y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

all libraries used

import numpy as np
import pandas as pd
import statsmodels.api as sm
import seaborn as sns
import matplotlib.pyplot as plt
from sklearn.preprocessing import scale, StandardScaler
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report
from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.neural_network import MLPClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import GradientBoostingClassifier

import warnings
warnings.filterwarnings("ignore", category= DeprecationWarning)
warnings.filterwarnings("ignore", category= FutureWarning)

Logistic Regression

libraries used in this field

from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")

data.head()

	Pregnancies	Glucose	BloodPressure	SkinThickness	Insulin	BMI	DiabetesPedigreeFunction	Age	Outcome
0	6	148	72	35	0	33.6	0.627	50	1
1	1	85	66	29	0	26.6	0.351	31	0
2	8	183	64	0	0	23.3	0.672	32	1
3	1	89	66	23	94	28.1	0.167	21	0
4	0	137	40	35	168	43.1	2.288	33	1

data["Outcome"].value_counts()

Outcome
0    500
1    268
Name: count, dtype: int64

data.describe().T

	count	mean	std	min	25%	50%	75%	max
Pregnancies	768.0	3.845052	3.369578	0.000	1.00000	3.0000	6.00000	17.00
Glucose	768.0	120.894531	31.972618	0.000	99.00000	117.0000	140.25000	199.00
BloodPressure	768.0	69.105469	19.355807	0.000	62.00000	72.0000	80.00000	122.00
SkinThickness	768.0	20.536458	15.952218	0.000	0.00000	23.0000	32.00000	99.00
Insulin	768.0	79.799479	115.244002	0.000	0.00000	30.5000	127.25000	846.00
BMI	768.0	31.992578	7.884160	0.000	27.30000	32.0000	36.60000	67.10
DiabetesPedigreeFunction	768.0	0.471876	0.331329	0.078	0.24375	0.3725	0.62625	2.42
Age	768.0	33.240885	11.760232	21.000	24.00000	29.0000	41.00000	81.00
Outcome	768.0	0.348958	0.476951	0.000	0.00000	0.0000	1.00000	1.00

y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)

y.head()

0    1
1    0
2    1
3    0
4    1
Name: Outcome, dtype: int64

x.head()

	Pregnancies	Glucose	BloodPressure	SkinThickness	Insulin	BMI	DiabetesPedigreeFunction	Age
0	6	148	72	35	0	33.6	0.627	50
1	1	85	66	29	0	26.6	0.351	31
2	8	183	64	0	0	23.3	0.672	32
3	1	89	66	23	94	28.1	0.167	21
4	0	137	40	35	168	43.1	2.288	33

loj_model = LogisticRegression(solver="liblinear").fit(x,y)

loj_model.intercept_

array([-5.88679617])

loj_model.coef_

array([[ 1.16994476e-01,  2.83733435e-02, -1.68981359e-02,
         7.55145090e-04, -6.41407258e-04,  5.97201268e-02,
         6.76128123e-01,  7.23498971e-03]])

loj_model.predict(x)[0:10]

array([1, 0, 1, 0, 1, 0, 0, 1, 1, 0], dtype=int64)

y[0:10]

0    1
1    0
2    1
3    0
4    1
5    0
6    1
7    0
8    1
9    1
Name: Outcome, dtype: int64

y_pred= loj_model.predict(x)

confusion_matrix(y,y_pred) #karmaşıklık matriksi

array([[448,  52],
       [121, 147]], dtype=int64)

accuracy_score(y,y_pred) #doğruluk oranı

0.7747395833333334

print(classification_report(y,y_pred))

              precision    recall  f1-score   support

           0       0.79      0.90      0.84       500
           1       0.74      0.55      0.63       268

    accuracy                           0.77       768
   macro avg       0.76      0.72      0.73       768
weighted avg       0.77      0.77      0.77       768

loj_model.predict_proba(x)[0:10]

array([[0.3505852 , 0.6494148 ],
       [0.91692518, 0.08307482],
       [0.22489628, 0.77510372],
       [0.92127453, 0.07872547],
       [0.16759435, 0.83240565],
       [0.79886109, 0.20113891],
       [0.8800353 , 0.1199647 ],
       [0.27795677, 0.72204323],
       [0.32053464, 0.67946536],
       [0.92264521, 0.07735479]])

logit_roc_auc = roc_auc_score(y, loj_model.predict(x))

fpr, tpr, thresholds = roc_curve(y, loj_model.predict_proba(x) [:,1])

plt.figure()

plt.plot(fpr, tpr, label='AUC (area 1 %0.2f)' % logit_roc_auc)

plt.plot([0, 1], [0, 1],'r--')

plt.xlim([0.0, 1.0])

plt.ylim([0.0, 1.05])

plt.xlabel('False Positive Rate')

plt.ylabel('True Positive Rate')

plt.title('Receiver operating characteristic')

plt.legend(loc="lower right")

plt.savefig('Log_ROC')

plt.show()

x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

loj_model= LogisticRegression(solver= "liblinear").fit(x_train,y_train)

y_pred= loj_model.predict(x_test)

print(accuracy_score(y_test,y_pred))

0.7532467532467533

cross_val_score(loj_model, x_test, y_test, cv= 10).mean()

0.7704710144927536

K-EN Yakın Komşu

libraries used in this field

from sklearn.neighbors import KNeighborsClassifier
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")

y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

knn_model= KNeighborsClassifier().fit(x_train,y_train)

knn_model

KNeighborsClassifier()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

y_pred= knn_model.predict(x_test)

accuracy_score(y_test,y_pred)

0.6883116883116883

print(classification_report(y_test,y_pred))

              precision    recall  f1-score   support

           0       0.77      0.75      0.76       151
           1       0.55      0.56      0.56        80

    accuracy                           0.69       231
   macro avg       0.66      0.66      0.66       231
weighted avg       0.69      0.69      0.69       231

knn= KNeighborsClassifier()

knn_params= {"n_neighbors":np.arange(1,50)}

knn_cv_model= GridSearchCV(knn,knn_params, cv=10).fit(x_train,y_train)

knn_cv_model.best_score_

0.748637316561845

knn_cv_model.best_params_

{'n_neighbors': 11}

knn_tuned = KNeighborsClassifier(n_neighbors= 11).fit(x_train,y_train)

y_pred= knn_tuned.predict(x_test)

accuracy_score(y_test, y_pred)

0.7316017316017316

knn_tuned.score(x_test,y_test) # kısa yol

0.7316017316017316

Support Vector Machines (SVM)

Purpose: to separate the two classes we have

libraries used in this field

from sklearn.svm import SVC
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")

y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

svm_model= SVC(kernel= "linear").fit(x_train,y_train)

svm_model

SVC(kernel='linear')

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

y_pred= svm_model.predict(x_test)

accuracy_score(y_test,y_pred)

0.7445887445887446

svm_params= {"C":np.arange(1,5),
            "kernel":["linear","rbf"]}

svm= SVC()

svm_cv_model= GridSearchCV(svm,svm_params,cv=5,n_jobs=-1,verbose=2).fit(x_train,y_train)

Fitting 5 folds for each of 8 candidates, totalling 40 fits

svm_cv_model.best_score_

0.7839044652128765

svm_cv_model.best_params_

{'C': 2, 'kernel': 'linear'}

svm_tuned= SVC(C= 2, kernel= "linear").fit(x_train,y_train)

y_pred= svm_tuned.predict(x_test)

accuracy_score(y_test,y_pred)

0.7445887445887446

Artificial neural networks

libraries used in this field

from sklearn.neural_network import MLPClassifier
from sklearn.preprocessing import scale, StandardScaler
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")
y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

yapay sinir ağları genelde homojen veri setleri üzerinde daha iyi çalıştığından dolayı standartlaşma işlemi uygulanması gerekiyor

scaler= StandardScaler()

scaler.fit(x_train)
x_train= scaler.transform(x_train)

scaler.fit(x_test)
x_test= scaler.transform(x_test)

mlpc_model= MLPClassifier().fit(x_train,y_train)

mlpc_model

MLPClassifier()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

mlpc_model.coefs_

[array([[ 6.45444925e-02, -1.45724306e-01, -1.78402760e-03,
         -8.29426816e-02, -3.62620725e-02, -1.59528463e-01,
         -1.44318684e-01, -1.49082050e-01, -1.95805489e-01,
         -3.06053930e-02, -1.53487557e-01,  7.90392358e-02,
          1.25638476e-01,  3.84571164e-02,  1.70108154e-01,
          1.41215377e-02, -2.47106870e-02, -3.53728864e-02,
          1.00505266e-01, -6.55616563e-02,  2.26623514e-01,
          1.87855944e-01,  1.93569669e-04,  7.35469381e-02,
          1.57274965e-01,  1.01644594e-01, -1.63297986e-03,
         -8.19642063e-03, -2.53207312e-01, -4.61899481e-02,
          2.14739196e-01,  1.83492423e-01,  8.00502703e-04,
         -1.39858361e-01, -6.16501050e-02,  1.34984348e-02,
         -2.23150209e-01, -1.19055179e-01,  9.28647591e-03,
          1.43016576e-01,  1.52975700e-01,  3.51834441e-01,
          5.73214182e-02,  2.49236636e-01, -1.49158944e-02,
          1.52101575e-01, -2.50990382e-01,  4.97402923e-02,
          1.28931097e-01,  1.04352922e-02, -8.80367767e-02,
         -3.01753408e-02, -1.83668219e-01,  4.33561608e-12,
          1.27202928e-02,  1.45349208e-01,  7.79164903e-02,
          2.08299581e-01, -1.58498256e-01,  2.27995593e-02,
         -6.87691299e-02, -2.38446603e-01,  1.69554782e-01,
          1.50499144e-01, -2.22338183e-01,  1.23250273e-03,
          1.59213729e-01, -1.18626280e-01,  1.33341092e-01,
         -3.11806465e-02,  1.83037349e-01,  1.02779819e-02,
         -3.86333242e-02,  1.24532902e-05, -2.35207272e-04,
         -5.51430161e-02, -1.70419443e-02,  2.14958268e-02,
          2.70176692e-01, -1.97683530e-02,  1.41223781e-01,
         -8.36485874e-02, -2.99498495e-01, -6.12811184e-02,
          2.97456337e-01,  1.24946604e-02,  1.69774591e-02,
          1.99947558e-01,  6.74548596e-02, -1.94405399e-02,
         -8.49355756e-02, -9.48789814e-02, -1.88779987e-01,
          5.55960720e-02,  5.13421427e-10, -4.89607850e-05,
         -2.16935780e-02, -3.81318255e-03, -8.46394196e-02,
          1.39662624e-01],
        [-6.17429506e-02,  1.89932324e-01, -9.92554384e-03,
          3.00727310e-03,  7.85993375e-02,  1.73876586e-01,
          1.27155128e-01,  1.13811070e-01,  4.30577966e-02,
          1.30710371e-01,  1.23900717e-01,  2.04818148e-01,
         -1.53514809e-01,  8.36626090e-02,  1.28983069e-01,
         -6.64380846e-02,  1.13257066e-01, -4.16908534e-02,
          2.18001567e-01, -1.97507694e-01,  1.63362075e-01,
          8.32404644e-02, -1.19162227e-03, -2.79011850e-02,
          1.49146606e-01, -1.60261161e-01, -1.97690932e-03,
          1.08454914e-01,  1.22357305e-01,  4.02657446e-04,
          6.44963045e-02, -1.02895558e-01, -4.38818799e-02,
          1.50243271e-01,  1.61847820e-01, -8.43676086e-02,
         -9.96789612e-02, -8.68465742e-02, -2.55412110e-02,
          1.45803201e-01,  1.04112291e-01, -2.61393714e-01,
          1.15361899e-01, -1.08366608e-01,  9.45407045e-02,
          1.36953441e-01, -4.26357275e-02, -1.70022706e-01,
          1.02092604e-01,  1.17586815e-01, -2.30035034e-01,
         -1.88930419e-01,  2.60815279e-02, -6.94212143e-11,
          1.21222854e-01, -1.31665234e-01,  6.08379031e-03,
          1.05211796e-01,  1.58213198e-01, -1.47996291e-01,
         -7.26469915e-03,  9.02698405e-02,  3.27559640e-02,
         -1.34424919e-01,  1.09899344e-01, -2.39671094e-03,
          1.90033301e-02, -3.30784244e-02,  2.30338438e-01,
         -2.80921429e-02, -7.24766002e-02,  1.53925216e-01,
          1.85560104e-01, -1.40241439e-03, -2.73889705e-03,
          2.88868322e-03, -2.07841614e-02,  4.15724177e-02,
          4.41319293e-02,  1.82312249e-01,  5.35194926e-03,
          1.23306694e-01, -5.96764385e-02,  1.57219179e-01,
         -6.05453269e-02,  9.16376411e-02, -1.11127826e-01,
         -2.14891899e-01,  1.35341786e-01,  5.72363580e-02,
          1.03645078e-01,  8.17621140e-02,  5.09157258e-02,
          4.48876460e-02, -1.19523099e-02, -2.07936512e-04,
          6.34009706e-02, -8.54078858e-02, -5.10941414e-02,
          1.10010215e-01],
        [ 2.16250578e-02,  3.85475486e-02, -8.79905014e-03,
         -2.53978534e-02, -5.31461567e-02,  1.81096568e-01,
          1.09268437e-01, -1.25524046e-01,  1.55088796e-01,
          1.47499977e-02, -9.78400983e-02,  1.32408180e-01,
         -3.12269275e-02, -2.58302512e-03, -6.29223030e-03,
          1.98099332e-01, -1.33905347e-01, -3.70129012e-02,
         -8.48295065e-02,  1.81249773e-01,  2.27901993e-01,
         -1.91867559e-01, -1.08718435e-02,  3.22704678e-02,
          7.35042466e-02,  9.20512452e-02, -9.70290080e-04,
          1.05985060e-01,  5.40076104e-02,  1.35986858e-01,
          4.62924428e-02,  4.62988276e-02,  5.12968058e-02,
         -1.58904051e-01, -8.22076502e-02, -1.68601992e-01,
         -5.72256296e-02, -1.82413626e-01,  2.32591055e-01,
          1.14054752e-01,  2.18622560e-01, -1.29986990e-01,
          1.03217300e-01,  4.69308914e-02,  2.00663360e-01,
          1.29062856e-01,  1.88247946e-01, -2.56112156e-02,
         -1.92117808e-02,  2.19725869e-01,  1.62743839e-01,
         -3.10413117e-02,  1.23333714e-01, -3.36596565e-03,
         -1.60354164e-01,  1.57046500e-01, -3.94233278e-02,
          1.05390077e-01, -5.15769323e-02,  9.99720582e-02,
         -8.49626279e-02, -5.40765147e-02,  2.35437914e-01,
          2.30835336e-01, -1.43484262e-01, -9.39181458e-03,
          7.30024143e-02,  1.38173126e-01, -1.41544885e-01,
         -1.81211653e-02,  3.77297197e-02,  1.34435442e-01,
         -2.13866093e-01, -3.97548365e-03, -1.52875935e-02,
          2.06732229e-01, -1.75162050e-02,  2.75646757e-02,
         -2.61778929e-02,  7.17691012e-02,  4.95131970e-02,
         -1.15188872e-01, -9.30634090e-02, -2.63505989e-01,
         -4.25628972e-02,  1.80000754e-01, -2.48665206e-01,
         -1.11366588e-02,  2.29619705e-01,  1.05304060e-02,
          1.55541653e-01, -2.13023892e-01, -1.82616829e-01,
          8.53990814e-02, -1.40141640e-04,  1.91837504e-05,
         -1.38056107e-01, -9.72921586e-02, -2.12988790e-01,
         -1.84146937e-01],
        [ 1.82410391e-01, -1.57341988e-01, -1.52453967e-02,
          1.85398170e-01, -9.02732215e-02, -1.04937795e-01,
          2.31750731e-01, -1.41412500e-01,  1.72655504e-01,
          1.36059229e-01,  1.19023774e-01, -3.62965779e-02,
          1.59538881e-01,  2.17471130e-01, -1.46106420e-01,
         -8.73155750e-02,  9.65726115e-02, -1.89004637e-02,
          2.27983973e-01, -1.20496641e-01,  2.06237775e-01,
          1.02285890e-01, -1.28699310e-04, -1.31761096e-01,
         -1.32998988e-01,  1.76367386e-02,  3.19893607e-09,
          1.90992744e-01,  9.23757742e-02, -2.31303476e-01,
          7.31910669e-02, -2.31047167e-01, -6.39094393e-02,
         -1.50648951e-01,  4.64727316e-02, -1.12335002e-01,
         -1.43127712e-01, -1.96547763e-02, -9.02834485e-02,
          3.37964241e-02, -1.83374326e-01, -2.97640351e-03,
         -1.54890066e-01,  2.73686932e-01,  7.93513525e-02,
          1.33671655e-01,  2.28587267e-01,  1.21366669e-01,
         -2.58258078e-02,  1.92242587e-01, -8.24420283e-02,
          2.35770083e-02,  1.38706850e-01,  1.08319399e-06,
          1.69465380e-01,  1.23490351e-02,  1.54306023e-01,
          1.86409321e-01, -1.57301441e-01,  2.17101384e-01,
         -1.75613497e-01,  9.66944905e-02, -1.76739533e-01,
          9.08781234e-02, -1.30773175e-01, -8.24090349e-04,
          2.23363166e-01, -1.94929369e-01, -2.40538198e-01,
         -1.34702933e-02,  3.87011395e-03,  2.69169982e-02,
         -2.07556188e-01,  4.23299640e-06,  6.42001824e-05,
          4.89505807e-02, -8.34726151e-03,  1.61338477e-02,
          1.31669852e-01,  2.45683871e-02, -2.13537566e-01,
         -4.98340304e-02, -2.28718899e-02, -4.08941379e-02,
         -8.22232895e-03, -7.49184636e-02,  9.01000141e-03,
          2.62588938e-02, -1.34374966e-01,  1.66672524e-01,
          1.56894886e-01,  1.39015270e-01, -2.90355950e-02,
          7.75201470e-03, -8.19200474e-03,  3.99920570e-05,
         -1.24367053e-01, -4.52664597e-02, -8.29741936e-02,
         -2.06740553e-01],
        [-1.25976859e-01, -3.14514979e-02, -1.75449139e-03,
         -5.57224183e-03,  2.39350174e-01, -1.06871921e-01,
         -2.09548581e-01, -1.31939652e-01,  1.80856378e-01,
         -9.89520194e-02,  8.28924521e-02,  1.97004632e-01,
         -1.95619669e-01, -2.45212570e-01, -9.66723975e-02,
         -1.45359946e-01, -2.27272565e-01,  2.78246238e-10,
         -3.83298879e-02,  5.63257776e-02,  1.77252523e-01,
         -3.34766009e-01, -2.55252850e-03,  1.74162057e-01,
          3.06041599e-02,  1.72546510e-01,  1.54385693e-07,
         -2.52142478e-01,  1.18476655e-01, -1.82406166e-01,
         -9.08812660e-02,  1.33267274e-02,  9.41680871e-02,
          5.33205074e-02,  1.56577619e-01, -2.57938989e-03,
          1.75930285e-01,  1.21396434e-01,  3.99627008e-02,
          8.28631392e-02, -9.49519188e-02, -1.60880890e-01,
         -7.17937284e-02, -2.38681906e-01, -1.05976938e-01,
         -2.47267880e-01,  1.93407779e-01,  1.94482391e-01,
          4.06305109e-02, -5.79530342e-02,  9.97977664e-02,
          1.22402365e-01, -1.41395073e-01, -4.67030339e-03,
         -1.14302885e-01, -1.86860051e-01, -1.68928169e-01,
          8.21642208e-02,  2.23881022e-01,  1.01704870e-02,
          1.75984877e-01, -3.91999757e-02, -2.54469873e-02,
          2.15365767e-01, -2.17424443e-01, -3.79154306e-05,
          1.85325418e-01, -9.24462617e-02,  3.11143173e-02,
         -7.89052045e-03, -1.43292960e-05, -1.05032714e-01,
          4.80345567e-02, -1.28326563e-03, -6.95243224e-03,
         -1.54740927e-01,  5.14063381e-10,  1.05338262e-01,
          1.12836180e-01, -1.64037381e-01, -6.48217177e-02,
         -1.00283118e-01,  2.32821509e-01, -3.60428979e-02,
         -5.54232828e-04, -1.54984539e-01,  2.17591882e-01,
          1.35645588e-01, -6.53821049e-02,  1.27073928e-01,
          1.83049364e-01,  1.57740863e-01, -1.93747071e-01,
          3.68506374e-03, -1.79799669e-05, -3.16284934e-06,
          1.16073461e-01,  1.66318818e-01,  1.89982209e-01,
          1.00001660e-01],
        [ 1.74587877e-02,  1.78527962e-01,  1.49599683e-03,
         -1.74046282e-01, -1.33874487e-01, -1.60634321e-01,
         -1.28921435e-01,  1.35547218e-01,  1.85346464e-01,
         -1.62144799e-01, -2.03648916e-01,  6.35606703e-02,
          1.96461178e-01, -2.16655487e-01,  2.30884161e-01,
          8.44688652e-02,  2.35630668e-01,  2.77914963e-02,
          3.05852062e-02,  1.79400312e-01, -1.54066473e-01,
          2.10278345e-01, -8.02921236e-03, -7.01168749e-02,
          1.02591610e-01,  2.08532059e-01, -7.73018455e-04,
          1.67387949e-01,  4.43725252e-02,  1.62756951e-01,
          7.93096113e-02, -1.08090669e-01,  8.52141203e-02,
          2.25017239e-01, -1.08406849e-01,  7.17143942e-02,
          3.73078261e-02,  9.92099255e-02,  2.08838180e-01,
          3.77133860e-02, -1.42946507e-01,  2.24231375e-01,
          1.18292067e-01,  1.25443716e-01,  1.54470393e-01,
         -2.12375875e-01, -3.64698518e-02, -2.19679249e-01,
          3.61611932e-02, -1.42750166e-01,  5.60233157e-02,
          1.21175689e-01,  7.22978716e-02, -5.65260198e-04,
         -3.36038092e-03,  2.08138022e-01, -1.23052629e-01,
         -1.07585462e-01, -3.77746738e-02, -2.06915696e-01,
          1.59512053e-01, -1.18510759e-01, -9.69541921e-02,
          1.67477196e-02,  1.11330368e-01, -1.20369740e-02,
         -7.95784047e-04, -7.14756996e-03,  1.33358116e-01,
         -2.29615954e-02, -1.83186892e-01, -7.10701260e-02,
          2.91711528e-02, -1.32263368e-04,  5.83452094e-05,
         -2.13006111e-01, -1.75136878e-02, -1.18844971e-01,
          6.45613342e-02,  9.53366310e-02,  1.30249744e-01,
         -1.32345394e-01,  2.00611981e-03, -1.30884295e-01,
          8.04802550e-02,  1.11115101e-01,  1.42200541e-01,
          1.15293165e-01, -4.00743052e-03, -1.42283419e-01,
         -1.84300444e-01,  6.91776300e-02,  1.06590000e-01,
         -5.85449793e-02, -9.99517778e-04, -4.80292072e-03,
         -2.11624030e-01,  1.46322987e-01, -9.73446054e-02,
         -4.46520018e-02],
        [ 7.22393730e-02, -1.88553498e-01, -7.94033989e-03,
         -1.05917911e-01, -8.88747307e-03,  2.13844754e-01,
          3.57707795e-02,  2.68590464e-01,  1.39424668e-02,
         -9.47142050e-03,  5.73284016e-03, -6.02683667e-02,
          5.59174437e-02,  1.95761125e-01, -4.70383033e-02,
         -1.58122471e-02,  1.88011640e-01,  1.06568288e-02,
         -2.04762309e-01,  1.08804394e-01,  1.22023815e-02,
          1.60200005e-02,  4.70004286e-10,  8.17144354e-02,
         -4.08214856e-02, -1.29112868e-01,  3.37455282e-08,
          9.60875995e-02,  1.40195747e-01,  4.62906471e-02,
         -1.64748273e-01,  2.11700183e-01, -2.29922985e-01,
          1.46063583e-01, -1.37174477e-01, -2.03700360e-01,
         -7.39758494e-02,  1.70556762e-01,  1.99564556e-01,
         -2.52253878e-02, -1.61206276e-01, -8.95778074e-02,
         -2.13604547e-02,  4.70364117e-01,  6.73413436e-02,
         -1.29587377e-01,  1.34013717e-01, -4.84177194e-02,
         -2.06247148e-01,  2.03789106e-01, -2.24239399e-01,
         -8.28956793e-02, -9.25723769e-02, -1.40985282e-04,
         -5.05824744e-02, -1.73807961e-01,  4.29571466e-02,
          6.07576678e-02,  1.37893840e-02,  1.03588716e-01,
         -2.10123769e-01, -1.47280147e-01,  5.20382721e-02,
         -1.23459719e-01,  1.53881850e-01, -2.67547013e-04,
          6.83981626e-02,  8.16296529e-02,  1.00922403e-01,
          6.59983796e-03, -1.27866050e-03, -9.14148254e-03,
          2.64823496e-02, -4.38874527e-11, -9.04063914e-03,
         -2.12350612e-01, -1.68270576e-02, -2.06101989e-01,
          1.10342659e-01, -8.10634124e-02,  3.21737446e-02,
          2.40380162e-01,  9.56745202e-02,  2.31795154e-01,
          1.24466449e-01,  1.55592296e-01,  1.47093384e-01,
          4.18839510e-02,  1.52291462e-01, -7.49036879e-02,
         -1.78949129e-01, -7.29837866e-02,  3.17530968e-02,
          1.63135576e-01,  5.45616230e-03, -5.14314002e-10,
          6.22721378e-02,  1.34177188e-01, -6.90255532e-02,
          1.39537375e-01],
        [ 3.08104958e-02, -1.44037531e-01, -3.06927278e-12,
          5.34702447e-02, -2.89297084e-02, -2.98064743e-02,
          1.58779479e-01,  3.91920206e-02, -3.45644611e-02,
          1.86828455e-01, -1.22037055e-01, -1.87045031e-02,
          6.01835353e-02,  2.29396566e-02, -1.32325054e-01,
         -4.78617960e-02,  2.06928311e-01,  6.17921287e-03,
          1.46844796e-01,  4.27299436e-03,  4.53881010e-02,
          5.25282517e-02, -7.44972548e-03, -5.94706199e-02,
          8.98869251e-02, -1.81493977e-01, -3.11495119e-04,
          7.94565885e-02,  4.46743043e-02, -1.68734459e-01,
         -2.05193004e-01,  7.21662820e-02,  9.06218762e-02,
         -3.19823109e-02,  6.58798037e-02,  1.17684682e-01,
         -5.52421585e-02,  1.67094380e-01, -1.98949693e-01,
         -9.60967729e-03, -1.17091827e-01,  2.07215674e-01,
         -1.61748207e-01, -3.94288325e-02,  1.87538462e-01,
          1.28384993e-02, -1.35759379e-01,  2.14628214e-01,
          1.69129211e-01, -2.07669695e-01,  1.11300672e-01,
         -2.69744120e-01, -1.26641907e-03, -8.75268807e-03,
          1.43950569e-01,  9.66290682e-02,  4.38652406e-02,
          2.11554415e-01,  1.01274862e-01,  8.21879693e-02,
         -1.23968104e-01,  1.23289404e-01, -5.26739475e-02,
          1.70099312e-01,  1.45440496e-01, -6.56505043e-03,
         -3.67608177e-02,  1.94719127e-01, -1.97071658e-01,
         -1.05803814e-02,  6.03680968e-02, -1.14625838e-01,
          1.91643765e-01,  5.53451850e-04,  1.24270509e-03,
         -2.08386456e-01, -1.85177584e-02, -1.89299820e-01,
         -1.51554768e-01, -9.80680866e-03, -1.31688828e-02,
         -2.30232536e-01, -6.02653102e-02,  3.71492361e-02,
          2.24249576e-02, -1.56879010e-01, -1.38809227e-01,
          2.13641109e-01, -9.77098473e-03, -1.47123180e-01,
          1.65972481e-01,  9.90208099e-02, -1.87630307e-01,
          7.93925546e-02, -1.19630478e-02, -1.07804154e-04,
         -1.40597717e-01, -2.17739956e-01, -1.45209940e-01,
         -1.54306522e-01]]),
 array([[ 2.68942122e-01],
        [-4.77370667e-02],
        [-7.04462921e-05],
        [-3.48637116e-01],
        [-1.54190190e-01],
        [ 1.56047661e-01],
        [-1.15128934e-01],
        [-2.88210932e-02],
        [ 7.87701257e-02],
        [ 1.47691390e-01],
        [ 2.01749746e-01],
        [-7.88019764e-02],
        [-3.88967464e-02],
        [ 1.42715936e-01],
        [-4.63054725e-02],
        [-9.99356262e-02],
        [-7.66369561e-02],
        [ 3.10586185e-02],
        [-2.35599221e-01],
        [-1.06107323e-01],
        [ 1.87552133e-01],
        [ 6.24427995e-02],
        [-1.21028705e-04],
        [-2.17999601e-04],
        [ 1.44869813e-01],
        [-1.26987162e-02],
        [-1.02012937e-02],
        [ 7.98791916e-02],
        [-7.05093504e-02],
        [ 1.79744820e-01],
        [ 2.87107107e-02],
        [ 1.43051771e-01],
        [ 5.58333797e-02],
        [-1.10246211e-01],
        [ 2.24089520e-01],
        [ 1.38007383e-01],
        [-2.50835681e-01],
        [ 5.61972007e-02],
        [-1.94017230e-01],
        [ 2.12647951e-01],
        [ 6.05501156e-02],
        [ 1.91510545e-01],
        [-1.21138397e-01],
        [ 1.19052801e-01],
        [ 8.24324866e-02],
        [-1.28314426e-01],
        [-1.46003329e-01],
        [ 7.75797680e-02],
        [-2.19151116e-01],
        [-7.21864873e-02],
        [ 1.94122901e-01],
        [-8.77937828e-02],
        [ 1.94179217e-01],
        [-2.39947580e-11],
        [ 8.33812027e-02],
        [ 3.59993859e-02],
        [-9.39927185e-02],
        [ 4.67064574e-02],
        [ 1.57162002e-01],
        [-1.32042312e-01],
        [ 1.16541501e-01],
        [-1.49032028e-01],
        [-1.53736868e-01],
        [-1.68269540e-01],
        [ 1.70568474e-01],
        [ 1.35753686e-02],
        [-2.91458401e-02],
        [-1.26415369e-01],
        [-9.61228077e-03],
        [ 1.11587511e-02],
        [-2.45383693e-02],
        [-3.87288232e-02],
        [-1.78066339e-01],
        [-7.50446016e-04],
        [-1.15989931e-02],
        [-1.21652299e-01],
        [-7.68644733e-03],
        [ 2.12088482e-01],
        [ 8.53791564e-02],
        [-4.52203282e-02],
        [ 2.02148051e-01],
        [ 2.03592289e-01],
        [-9.52405747e-02],
        [-1.99828933e-01],
        [ 3.13333247e-01],
        [-1.44388008e-01],
        [ 3.84985337e-02],
        [ 1.16026415e-01],
        [ 1.84456113e-01],
        [-1.43387592e-01],
        [-2.45659576e-01],
        [ 2.01989796e-01],
        [ 1.59320723e-01],
        [-7.88899370e-02],
        [ 1.03192357e-07],
        [-1.09601807e-03],
        [-7.45593973e-02],
        [ 8.62688225e-02],
        [-4.49390354e-02],
        [-3.09125248e-02]])]

?mlpc_model

y_pred= mlpc_model.predict(x_test)

accuracy_score(y_test,y_pred)

0.7316017316017316

mlpc_params= {"hidden_layer_sizes":[(10,10),(100,100,100),(100,100),(3,5)],
             "alpha":[1,5,0.1,0.01,0.03,0.05,0.0001]}

mlpc= MLPClassifier(solver="lbfgs",activation= "logistic") # sınıflandırma problemlerinde activationu logistic yapmamız gerekiyor 

mlpc_cv_model= GridSearchCV(mlpc,mlpc_params,cv=10,n_jobs=-1,verbose=2).fit(x_train,y_train)

Fitting 10 folds for each of 28 candidates, totalling 280 fits

C:\Users\mehdi\anaconda3\Lib\site-packages\sklearn\neural_network\_multilayer_perceptron.py:546: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.

Increase the number of iterations (max_iter) or scale the data as shown in:
    https://scikit-learn.org/stable/modules/preprocessing.html
  self.n_iter_ = _check_optimize_result("lbfgs", opt_res, self.max_iter)

mlpc_cv_model

GridSearchCV(cv=10,
             estimator=MLPClassifier(activation='logistic', solver='lbfgs'),
             n_jobs=-1,
             param_grid={'alpha': [1, 5, 0.1, 0.01, 0.03, 0.05, 0.0001],
                         'hidden_layer_sizes': [(10, 10), (100, 100, 100),
                                                (100, 100), (3, 5)]},
             verbose=2)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

mlpc_cv_model.best_params_

{'alpha': 5, 'hidden_layer_sizes': (100, 100)}

mlpc_tuned= MLPClassifier(solver="lbfgs",alpha=5,hidden_layer_sizes=(100,100),activation= "logistic").fit(x_train,y_train)

y_pred= mlpc_tuned.predict(x_test)

accuracy_score(y_test,y_pred)

0.7359307359307359

CART

libraries used in this field

from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")
y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42);

cart_model= DecisionTreeClassifier().fit(x_train,y_train)

cart_model

DecisionTreeClassifier()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

y_pred= cart_model.predict(x_test)

accuracy_score(y_test,y_pred)

0.7056277056277056

cart= DecisionTreeClassifier()

cart_params= {"max_depth": [1,3,5,8,10],
             "min_samples_split":[2,3,5,10,20,50]}

cart_cv_model= GridSearchCV(cart,cart_params,n_jobs=-1,verbose=2).fit(x_train,y_train)

Fitting 5 folds for each of 30 candidates, totalling 150 fits

cart_cv_model.best_params_

{'max_depth': 5, 'min_samples_split': 50}

cart_tuned= DecisionTreeClassifier(max_depth=5,min_samples_split=50).fit(x_train,y_train)

y_pred= cart_tuned.predict(x_test)

accuracy_score(y_test,y_pred)

0.7575757575757576

Random Forests

libraries used in this field

from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")
y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

rf_model= RandomForestClassifier().fit(x_train,y_train)

rf_model

RandomForestClassifier()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

y_pred= rf_model.predict(x_test)

accuracy_score(y_test, y_pred)

0.7532467532467533

rf= RandomForestClassifier()

rf_params= {"n_estimators":[100,200,500,1000],
           "max_features":[3,5,7,8],
           "min_samples_split": [2,5,10,20]}

rf_cv_model= GridSearchCV(rf,rf_params,cv=10,n_jobs=-1,verbose=2).fit(x_train,y_train)

Fitting 10 folds for each of 64 candidates, totalling 640 fits

rf_cv_model.best_params_

{'max_features': 5, 'min_samples_split': 5, 'n_estimators': 200}

rf_tuned= RandomForestClassifier(n_estimators=200,max_features=5,min_samples_split=5).fit(x_train,y_train)

y_pred= rf_tuned.predict(x_test)

accuracy_score(y_test,y_pred)

0.7575757575757576

# değişken önem düzeyi

rf_tuned.feature_importances_

array([0.05501306, 0.35020139, 0.07096524, 0.05014943, 0.0529208 ,
       0.17518745, 0.10316439, 0.14239823])

feature= pd.Series(rf_tuned.feature_importances_,
                 index=x_train.columns).sort_values(ascending=False)

sns.barplot(x=feature, y=feature.index)
plt.xlabel('Değişken Önem Skorlarrı')
plt.ylabel('Değişkenler')
plt.title('Değişken Önem Düzeyleri')
plt.show()

Gradient Boosting Machines (GBM)

libraries used in this field

from sklearn.ensemble import GradientBoostingClassifier
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")
y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

gbm_model= GradientBoostingClassifier().fit(x_train,y_train)

gbm_model

GradientBoostingClassifier()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

?gbm_model

y_pred= gbm_model.predict(x_test)

accuracy_score(y_test,y_pred)

0.7532467532467533

gbm= GradientBoostingClassifier()

gbm_params= {"learning_rate":[0.1,0.01,0.001,0.05],
            "n_estimators":[100,300,500,1000],
            "max_depth":[2,3,5,8]}

gbm_cv_model= GridSearchCV(gbm,gbm_params,cv=10,n_jobs=-1,verbose=2).fit(x_train,y_train)

gbm_cv_model.best_params_

gbm_tuned= GradientBoostingClassifier(learning_rate=0.01,
                                      n_estimators=5,
                                      max_depth=500).fit(x_train,y_train)

y_pred= gbm_tuned.predict(x_test)

accuracy_score(y_test,y_pred)

0.6536796536796536

# değişken önem düzeyi

feature= pd.Series(gbm_tuned.feature_importances_,
                 index=x_train.columns).sort_values(ascending=False)

sns.barplot(x=feature, y=feature.index)
plt.xlabel('Değişken Önem Skorlarrı')
plt.ylabel('Değişkenler')
plt.title('Değişken Önem Düzeyleri')
plt.show()

XGBoost

libraries used in this field

!pip insatll xgboost

ERROR: unknown command "insatll" - maybe you meant "install"

from xgboost import XGBClassifier
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")
y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

xgb_model= XGBClassifier().fit(x_train,y_train)

?xgb_model

y_pred=xgb_model.predict(x_test)

accuracy_score(y_test,y_pred)

0.7272727272727273

xgb= XGBClassifier()

xgb_params= {"n_estimators":[100,500,1000],
            "subsample":[0.6,0.8,1],
            "max_depth":[3,5,7],
            "learning_rate":[0.1,0.001,0.01]}

xgb_cv_model= GridSearchCV(xgb,xgb_params,cv=10,n_jobs=-1,verbose=2).fit(x_train,y_train)

Fitting 10 folds for each of 81 candidates, totalling 810 fits

xgb_cv_model.best_params_

{'learning_rate': 0.01, 'max_depth': 3, 'n_estimators': 500, 'subsample': 0.8}

xgb_tuned= XGBClassifier(learning_rate=0.01 ,
                        max_depth=3 ,
                        n_estimators=500 ,
                        subsample= 0.8).fit(x_train,y_train)

y_pred= xgb_tuned.predict(x_test)

accuracy_score(y_test,y_pred)

0.7619047619047619

feature= pd.Series(xgb_tuned.feature_importances_,
                 index=x_train.columns).sort_values(ascending=False)

sns.barplot(x=feature, y=feature.index)
plt.xlabel('Değişken Önem Skorlarrı')
plt.ylabel('Değişkenler')
plt.title('Değişken Önem Düzeyleri')
plt.show()

Light GBM

libraries used in this field

!pip insatll lightgbm

ERROR: unknown command "insatll" - maybe you meant "install"

from lightgbm import LGBMClassifier
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")
y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

lgbm_model= LGBMClassifier().fit(x_train,y_train)

[LightGBM] [Info] Number of positive: 188, number of negative: 349
[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.000526 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 606
[LightGBM] [Info] Number of data points in the train set: 537, number of used features: 8
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.350093 -> initscore=-0.618630
[LightGBM] [Info] Start training from score -0.618630
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf

lgbm_model

LGBMClassifier()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

?lgbm_model

y_predit= lgbm_model.predict(x_test)

accuracy_score(y_test,y_pred)

0.7619047619047619

lgbm= LGBMClassifier()

lgbm_params= {"learnin_rate":[0.001,0.01,0.01],
             "n_estimators":[200,500,100],
             "max_depth":[1,2,35,8]}

lgbm_cv_model= GridSearchCV(lgbm,lgbm_params,n_jobs=-1,verbose=2,cv=10).fit(x_train,y_train)

Fitting 10 folds for each of 36 candidates, totalling 360 fits
[LightGBM] [Warning] Unknown parameter: learnin_rate
[LightGBM] [Warning] Unknown parameter: learnin_rate
[LightGBM] [Info] Number of positive: 188, number of negative: 349
[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.000309 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 606
[LightGBM] [Info] Number of data points in the train set: 537, number of used features: 8
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.350093 -> initscore=-0.618630
[LightGBM] [Info] Start training from score -0.618630
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf

lgbm_cv_model.best_params_

{'learnin_rate': 0.001, 'max_depth': 2, 'n_estimators': 100}

lgbm_tuned= LGBMClassifier(learning_rate=0.01,
                          max_depth=1,
                          n_estimators=500).fit(x_train,y_train)

[LightGBM] [Info] Number of positive: 188, number of negative: 349
[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.000301 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 606
[LightGBM] [Info] Number of data points in the train set: 537, number of used features: 8
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.350093 -> initscore=-0.618630
[LightGBM] [Info] Start training from score -0.618630

y_pred= lgbm_tuned.predict(x_test)

accuracy_score(y_test,y_pred)

0.7575757575757576

feature= pd.Series(lgbm_tuned.feature_importances_,
                 index=x_train.columns).sort_values(ascending=False)

sns.barplot(x=feature, y=feature.index)
plt.xlabel('Değişken Önem Skorlarrı')
plt.ylabel('Değişkenler')
plt.title('Değişken Önem Düzeyleri')
plt.show()

Category Boosting (CatBoost)

libraries used in this field

!pip insatll catboost

ERROR: unknown command "insatll" - maybe you meant "install"

from catboost import CatBoostClassifier
from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, accuracy_score, mean_squared_error, r2_score, roc_auc_score, roc_curve, classification_report

data= pd.read_csv("C:\\data_set\\diabetes.csv")
y = data["Outcome"]
x= data.drop(["Outcome"],axis = 1)
x_train, x_test, y_train, y_test =train_test_split(x,y,test_size=0.30,random_state=42)

catb_model= CatBoostClassifier().fit(x_train,y_train,verbose=False)

y_pred= catb_model.predict(x_test)

accuracy_score(y_test,y_pred)

0.7402597402597403

catb= CatBoostClassifier(verbose=False)

catb_params= {"iterations":[200,500,1000],
             "learning_rate": [0.01,0.03,0.1],
             "depth": [4,5,8]}

# catb_cv_model= GridSearchCV(catb,catb_params,cv=5,n_jobs=-1,verbose=2).fit(x_train,y_train)

# catb_cv_model.best_params_

catb_tuned= CatBoostClassifier(depth=8,
                              iterations=200,
                              learning_rate=0.03).fit(x_train,y_train)

0:	learn: 0.6730317	total: 9.05ms	remaining: 1.8s
1:	learn: 0.6582932	total: 17ms	remaining: 1.68s
2:	learn: 0.6413374	total: 29.7ms	remaining: 1.95s
3:	learn: 0.6267884	total: 38.1ms	remaining: 1.87s
4:	learn: 0.6150338	total: 46.6ms	remaining: 1.82s
5:	learn: 0.5988569	total: 64.4ms	remaining: 2.08s
6:	learn: 0.5860830	total: 85.6ms	remaining: 2.36s
7:	learn: 0.5734710	total: 104ms	remaining: 2.5s
8:	learn: 0.5616034	total: 126ms	remaining: 2.67s
9:	learn: 0.5507407	total: 136ms	remaining: 2.59s
10:	learn: 0.5388119	total: 145ms	remaining: 2.49s
11:	learn: 0.5264274	total: 153ms	remaining: 2.4s
12:	learn: 0.5191433	total: 164ms	remaining: 2.35s
13:	learn: 0.5104236	total: 173ms	remaining: 2.3s
14:	learn: 0.5016788	total: 181ms	remaining: 2.23s
15:	learn: 0.4954697	total: 189ms	remaining: 2.18s
16:	learn: 0.4896572	total: 208ms	remaining: 2.23s
17:	learn: 0.4808284	total: 229ms	remaining: 2.31s
18:	learn: 0.4723731	total: 248ms	remaining: 2.37s
19:	learn: 0.4656953	total: 269ms	remaining: 2.42s
20:	learn: 0.4613141	total: 286ms	remaining: 2.44s
21:	learn: 0.4536779	total: 294ms	remaining: 2.38s
22:	learn: 0.4488597	total: 301ms	remaining: 2.32s
23:	learn: 0.4438188	total: 309ms	remaining: 2.26s
24:	learn: 0.4371743	total: 323ms	remaining: 2.26s
25:	learn: 0.4327707	total: 341ms	remaining: 2.28s
26:	learn: 0.4274522	total: 362ms	remaining: 2.32s
27:	learn: 0.4214705	total: 381ms	remaining: 2.34s
28:	learn: 0.4160825	total: 401ms	remaining: 2.36s
29:	learn: 0.4116521	total: 412ms	remaining: 2.33s
30:	learn: 0.4073682	total: 423ms	remaining: 2.3s
31:	learn: 0.4012053	total: 434ms	remaining: 2.28s
32:	learn: 0.3970894	total: 445ms	remaining: 2.25s
33:	learn: 0.3908936	total: 454ms	remaining: 2.21s
34:	learn: 0.3868677	total: 461ms	remaining: 2.17s
35:	learn: 0.3836666	total: 469ms	remaining: 2.14s
36:	learn: 0.3799969	total: 478ms	remaining: 2.1s
37:	learn: 0.3768569	total: 493ms	remaining: 2.1s
38:	learn: 0.3726554	total: 511ms	remaining: 2.11s
39:	learn: 0.3691778	total: 531ms	remaining: 2.12s
40:	learn: 0.3655939	total: 549ms	remaining: 2.13s
41:	learn: 0.3616921	total: 570ms	remaining: 2.14s
42:	learn: 0.3580472	total: 583ms	remaining: 2.13s
43:	learn: 0.3556024	total: 594ms	remaining: 2.1s
44:	learn: 0.3520438	total: 603ms	remaining: 2.08s
45:	learn: 0.3493385	total: 612ms	remaining: 2.05s
46:	learn: 0.3456439	total: 621ms	remaining: 2.02s
47:	learn: 0.3432385	total: 629ms	remaining: 1.99s
48:	learn: 0.3400849	total: 637ms	remaining: 1.96s
49:	learn: 0.3353309	total: 645ms	remaining: 1.94s
50:	learn: 0.3333735	total: 654ms	remaining: 1.91s
51:	learn: 0.3299440	total: 662ms	remaining: 1.88s
52:	learn: 0.3279711	total: 680ms	remaining: 1.89s
53:	learn: 0.3252567	total: 699ms	remaining: 1.89s
54:	learn: 0.3233230	total: 718ms	remaining: 1.89s
55:	learn: 0.3206790	total: 736ms	remaining: 1.89s
56:	learn: 0.3180869	total: 754ms	remaining: 1.89s
57:	learn: 0.3160875	total: 765ms	remaining: 1.87s
58:	learn: 0.3140489	total: 776ms	remaining: 1.85s
59:	learn: 0.3127124	total: 784ms	remaining: 1.83s
60:	learn: 0.3109795	total: 792ms	remaining: 1.8s
61:	learn: 0.3081993	total: 801ms	remaining: 1.78s
62:	learn: 0.3058535	total: 808ms	remaining: 1.76s
63:	learn: 0.3034111	total: 816ms	remaining: 1.73s
64:	learn: 0.3007186	total: 824ms	remaining: 1.71s
65:	learn: 0.2988902	total: 832ms	remaining: 1.69s
66:	learn: 0.2974293	total: 840ms	remaining: 1.67s
67:	learn: 0.2951830	total: 863ms	remaining: 1.67s
68:	learn: 0.2917267	total: 882ms	remaining: 1.67s
69:	learn: 0.2903607	total: 888ms	remaining: 1.65s
70:	learn: 0.2882471	total: 910ms	remaining: 1.65s
71:	learn: 0.2863627	total: 927ms	remaining: 1.65s
72:	learn: 0.2847011	total: 938ms	remaining: 1.63s
73:	learn: 0.2830881	total: 948ms	remaining: 1.61s
74:	learn: 0.2805163	total: 959ms	remaining: 1.6s
75:	learn: 0.2788447	total: 967ms	remaining: 1.58s
76:	learn: 0.2761505	total: 976ms	remaining: 1.56s
77:	learn: 0.2749231	total: 984ms	remaining: 1.54s
78:	learn: 0.2733597	total: 992ms	remaining: 1.52s
79:	learn: 0.2722927	total: 1000ms	remaining: 1.5s
80:	learn: 0.2709836	total: 1.01s	remaining: 1.48s
81:	learn: 0.2698701	total: 1.02s	remaining: 1.47s
82:	learn: 0.2683076	total: 1.04s	remaining: 1.47s
83:	learn: 0.2669109	total: 1.06s	remaining: 1.47s
84:	learn: 0.2647336	total: 1.08s	remaining: 1.46s
85:	learn: 0.2634293	total: 1.1s	remaining: 1.46s
86:	learn: 0.2621911	total: 1.11s	remaining: 1.44s
87:	learn: 0.2605138	total: 1.12s	remaining: 1.42s
88:	learn: 0.2594740	total: 1.13s	remaining: 1.4s
89:	learn: 0.2585352	total: 1.14s	remaining: 1.39s
90:	learn: 0.2569049	total: 1.16s	remaining: 1.38s
91:	learn: 0.2560202	total: 1.18s	remaining: 1.38s
92:	learn: 0.2543659	total: 1.2s	remaining: 1.38s
93:	learn: 0.2517455	total: 1.22s	remaining: 1.37s
94:	learn: 0.2506967	total: 1.23s	remaining: 1.36s
95:	learn: 0.2494585	total: 1.24s	remaining: 1.34s
96:	learn: 0.2473779	total: 1.25s	remaining: 1.33s
97:	learn: 0.2461214	total: 1.26s	remaining: 1.31s
98:	learn: 0.2449923	total: 1.27s	remaining: 1.29s
99:	learn: 0.2443683	total: 1.27s	remaining: 1.27s
100:	learn: 0.2436926	total: 1.28s	remaining: 1.26s
101:	learn: 0.2431051	total: 1.29s	remaining: 1.24s
102:	learn: 0.2418184	total: 1.3s	remaining: 1.22s
103:	learn: 0.2409622	total: 1.32s	remaining: 1.22s
104:	learn: 0.2403718	total: 1.34s	remaining: 1.21s
105:	learn: 0.2383218	total: 1.36s	remaining: 1.21s
106:	learn: 0.2374951	total: 1.38s	remaining: 1.2s
107:	learn: 0.2355808	total: 1.4s	remaining: 1.19s
108:	learn: 0.2347997	total: 1.41s	remaining: 1.18s
109:	learn: 0.2339094	total: 1.42s	remaining: 1.16s
110:	learn: 0.2323558	total: 1.43s	remaining: 1.14s
111:	learn: 0.2311310	total: 1.43s	remaining: 1.13s
112:	learn: 0.2300307	total: 1.44s	remaining: 1.11s
113:	learn: 0.2292299	total: 1.45s	remaining: 1.09s
114:	learn: 0.2286090	total: 1.46s	remaining: 1.08s
115:	learn: 0.2266665	total: 1.47s	remaining: 1.06s
116:	learn: 0.2256470	total: 1.48s	remaining: 1.05s
117:	learn: 0.2250197	total: 1.49s	remaining: 1.03s
118:	learn: 0.2242045	total: 1.51s	remaining: 1.02s
119:	learn: 0.2233188	total: 1.52s	remaining: 1.02s
120:	learn: 0.2222564	total: 1.54s	remaining: 1.01s
121:	learn: 0.2217455	total: 1.56s	remaining: 998ms
122:	learn: 0.2206488	total: 1.57s	remaining: 985ms
123:	learn: 0.2184009	total: 1.58s	remaining: 970ms
124:	learn: 0.2174793	total: 1.59s	remaining: 955ms
125:	learn: 0.2155648	total: 1.6s	remaining: 940ms
126:	learn: 0.2149142	total: 1.61s	remaining: 924ms
127:	learn: 0.2132945	total: 1.62s	remaining: 909ms
128:	learn: 0.2120564	total: 1.62s	remaining: 894ms
129:	learn: 0.2102406	total: 1.64s	remaining: 884ms
130:	learn: 0.2091022	total: 1.66s	remaining: 875ms
131:	learn: 0.2083464	total: 1.68s	remaining: 867ms
132:	learn: 0.2081746	total: 1.69s	remaining: 850ms
133:	learn: 0.2075415	total: 1.71s	remaining: 840ms
134:	learn: 0.2065361	total: 1.72s	remaining: 827ms
135:	learn: 0.2056145	total: 1.73s	remaining: 812ms
136:	learn: 0.2050039	total: 1.73s	remaining: 798ms
137:	learn: 0.2043998	total: 1.74s	remaining: 783ms
138:	learn: 0.2039071	total: 1.75s	remaining: 770ms
139:	learn: 0.2022818	total: 1.77s	remaining: 760ms
140:	learn: 0.2013156	total: 1.79s	remaining: 750ms
141:	learn: 0.2001301	total: 1.81s	remaining: 740ms
142:	learn: 0.1991370	total: 1.83s	remaining: 730ms
143:	learn: 0.1982389	total: 1.84s	remaining: 715ms
144:	learn: 0.1971621	total: 1.85s	remaining: 701ms
145:	learn: 0.1953999	total: 1.86s	remaining: 687ms
146:	learn: 0.1943254	total: 1.86s	remaining: 673ms
147:	learn: 0.1933229	total: 1.87s	remaining: 658ms
148:	learn: 0.1923080	total: 1.88s	remaining: 644ms
149:	learn: 0.1908673	total: 1.89s	remaining: 631ms
150:	learn: 0.1900431	total: 1.91s	remaining: 620ms
151:	learn: 0.1895706	total: 1.93s	remaining: 609ms
152:	learn: 0.1885998	total: 1.94s	remaining: 597ms
153:	learn: 0.1879384	total: 1.96s	remaining: 587ms
154:	learn: 0.1872544	total: 1.98s	remaining: 575ms
155:	learn: 0.1858131	total: 1.99s	remaining: 561ms
156:	learn: 0.1847201	total: 2s	remaining: 548ms
157:	learn: 0.1834393	total: 2.01s	remaining: 534ms
158:	learn: 0.1827037	total: 2.02s	remaining: 520ms
159:	learn: 0.1819670	total: 2.02s	remaining: 506ms
160:	learn: 0.1814510	total: 2.03s	remaining: 492ms
161:	learn: 0.1806783	total: 2.04s	remaining: 479ms
162:	learn: 0.1788436	total: 2.05s	remaining: 465ms
163:	learn: 0.1780870	total: 2.06s	remaining: 451ms
164:	learn: 0.1778217	total: 2.06s	remaining: 438ms
165:	learn: 0.1774635	total: 2.08s	remaining: 425ms
166:	learn: 0.1766113	total: 2.1s	remaining: 415ms
167:	learn: 0.1753497	total: 2.12s	remaining: 403ms
168:	learn: 0.1740266	total: 2.13s	remaining: 392ms
169:	learn: 0.1731058	total: 2.15s	remaining: 380ms
170:	learn: 0.1723178	total: 2.16s	remaining: 367ms
171:	learn: 0.1715993	total: 2.17s	remaining: 354ms
172:	learn: 0.1704719	total: 2.18s	remaining: 340ms
173:	learn: 0.1694361	total: 2.19s	remaining: 327ms
174:	learn: 0.1688269	total: 2.2s	remaining: 314ms
175:	learn: 0.1682591	total: 2.2s	remaining: 301ms
176:	learn: 0.1673707	total: 2.22s	remaining: 288ms
177:	learn: 0.1671943	total: 2.24s	remaining: 277ms
178:	learn: 0.1666458	total: 2.26s	remaining: 265ms
179:	learn: 0.1660139	total: 2.28s	remaining: 253ms
180:	learn: 0.1652719	total: 2.3s	remaining: 242ms
181:	learn: 0.1643594	total: 2.32s	remaining: 229ms
182:	learn: 0.1635202	total: 2.33s	remaining: 216ms
183:	learn: 0.1627135	total: 2.34s	remaining: 203ms
184:	learn: 0.1620158	total: 2.35s	remaining: 190ms
185:	learn: 0.1615713	total: 2.35s	remaining: 177ms
186:	learn: 0.1606615	total: 2.36s	remaining: 164ms
187:	learn: 0.1597244	total: 2.37s	remaining: 151ms
188:	learn: 0.1590019	total: 2.38s	remaining: 138ms
189:	learn: 0.1585646	total: 2.39s	remaining: 126ms
190:	learn: 0.1578091	total: 2.39s	remaining: 113ms
191:	learn: 0.1570803	total: 2.4s	remaining: 100ms
192:	learn: 0.1561039	total: 2.41s	remaining: 87.4ms
193:	learn: 0.1554807	total: 2.42s	remaining: 74.8ms
194:	learn: 0.1540806	total: 2.43s	remaining: 62.2ms
195:	learn: 0.1528199	total: 2.43s	remaining: 49.7ms
196:	learn: 0.1519209	total: 2.44s	remaining: 37.2ms
197:	learn: 0.1513105	total: 2.45s	remaining: 24.7ms
198:	learn: 0.1508391	total: 2.46s	remaining: 12.4ms
199:	learn: 0.1496336	total: 2.47s	remaining: 0us

y_pred= catb_tuned.predict(x_test)

accuracy_score(y_test,y_pred)

0.7489177489177489

feature= pd.Series(catb_tuned.feature_importances_,
                 index=x_train.columns).sort_values(ascending=False)

sns.barplot(x=feature, y=feature.index)
plt.xlabel('Değişken Önem Skorlarrı')
plt.ylabel('Değişkenler')
plt.title('Değişken Önem Düzeyleri')
plt.show()

Tüm Modellerin Karşılaştırılması

modeller =[
    knn_tuned,
    loj_model,
    svm_tuned,
    mlpc_tuned,
    cart_tuned,
    rf_tuned,
    gbm_tuned,
    catb_tuned,
    lgbm_tuned,
    xgb_tuned]
sonuc= []
sonuclar = pd.DataFrame(columns=["Modeller", "Accuracy"])

for model in modeller:
    isimler = model.__class__.__name__
    y_pred = model.predict(x_test)
    dogruluk = accuracy_score(y_test, y_pred)
    sonuc = pd.DataFrame([[isimler, dogruluk * 100]], columns=["Modeller", "Accuracy"])
    sonuclar = pd.concat([sonuclar, sonuc], ignore_index=True)

C:\Users\mehdi\anaconda3\Lib\site-packages\sklearn\base.py:464: UserWarning: X does not have valid feature names, but KNeighborsClassifier was fitted with feature names
  warnings.warn(
C:\Users\mehdi\anaconda3\Lib\site-packages\sklearn\base.py:464: UserWarning: X does not have valid feature names, but LogisticRegression was fitted with feature names
  warnings.warn(
C:\Users\mehdi\anaconda3\Lib\site-packages\sklearn\base.py:464: UserWarning: X does not have valid feature names, but SVC was fitted with feature names
  warnings.warn(
C:\Users\mehdi\anaconda3\Lib\site-packages\sklearn\base.py:464: UserWarning: X does not have valid feature names, but DecisionTreeClassifier was fitted with feature names
  warnings.warn(
C:\Users\mehdi\anaconda3\Lib\site-packages\sklearn\base.py:464: UserWarning: X does not have valid feature names, but RandomForestClassifier was fitted with feature names
  warnings.warn(
C:\Users\mehdi\anaconda3\Lib\site-packages\sklearn\base.py:464: UserWarning: X does not have valid feature names, but GradientBoostingClassifier was fitted with feature names
  warnings.warn(

sonuclar = sonuclar.sort_values(by='Accuracy', ascending=False)

sns.barplot(x='Accuracy',y='Modeller',data= sonuclar, color="r")
plt.xlabel('Accuracy %')
plt.title('Modellerin Doğruluk Oranları');

sonuclar

	Modeller	Accuracy
1	LogisticRegression	76.623377
9	XGBClassifier	76.190476
4	DecisionTreeClassifier	75.757576
8	LGBMClassifier	75.757576
7	CatBoostClassifier	74.891775
2	SVC	74.458874
5	RandomForestClassifier	74.025974
0	KNeighborsClassifier	73.160173
3	MLPClassifier	72.727273
6	GradientBoostingClassifier	65.367965

yüksek korelasyonu sahip değişkenler den bazıları çıkarılır çünkü benzer şeyler açıklamaktadırlar

?gbm_model

# Logistic Regression
loj_model = LogisticRegression(solver="liblinear").fit(x,y)

# K-EN Yakın Komşu
knn_model= KNeighborsClassifier().fit(x_train,y_train)
knn_params= {"n_neighbors":np.arange(1,50)}

# Support Vector Machines (SVM)
svm_model= SVC(kernel= "linear").fit(x_train,y_train)
svm_params= {"C":np.arange(1,5),
            "kernel":["linear","rbf"]}

# Artificial neural networks
scaler= StandardScaler()
scaler.fit(x_train)
x_train= scaler.transform(x_train)
scaler.fit(x_test)
x_test= scaler.transform(x_test)
mlpc_model= MLPClassifier().fit(x_train,y_train)
mlpc_params= {"hidden_layer_sizes":[(10,10),(100,100,100),(100,100),(3,5)],
             "alpha":[1,5,0.1,0.01,0.03,0.05,0.0001]}
mlpc= MLPClassifier(solver="lbfgs",activation= "logistic") # sınıflandırma problemlerinde activationu logistic yapmamız gerekiyor 


# CART
cart_model= DecisionTreeClassifier().fit(x_train,y_train)
cart_params= {"max_depth": [1,3,5,8,10],
             "min_samples_split":[2,3,5,10,20,50]}


# Random Forest
rf_model= RandomForestClassifier().fit(x_train,y_train)
rf_params= {"n_estimators":[100,200,500,1000],
           "max_features":[3,5,7,8],
           "min_samples_split": [2,5,10,20]}


# Gradient Boosting Machines (GBM)
gbm_model= GradientBoostingClassifier().fit(x_train,y_train)
gbm_params= {"learning_rate":[0.1,0.01,0.001,0.05],
            "n_estimators":[100,300,500,1000],
            "max_depth":[2,3,5,8]}

# XGBoost
xgb_model= XGBClassifier().fit(x_train,y_train)
xgb_params= {"n_estimators":[100,500,1000],
            "subsample":[0.6,0.8,1],
            "max_depth":[3,5,7],
            "learning_rate":[0.1,0.001,0.01]}

# Light GBM
lgbm_model= LGBMClassifier().fit(x_train,y_train)
lgbm_params= {"learnin_rate":[0.001,0.01,0.01],
             "n_estimators":[200,500,100],
             "max_depth":[1,2,35,8]}


# Category Boosting (CatBoost)
catb_model= CatBoostClassifier().fit(x_train,y_train,verbose=False)
catb_params= {"iterations":[200,500,1000],
             "learning_rate": [0.01,0.03,0.1],
             "depth": [4,5,8]}

C:\Users\mehdi\anaconda3\Lib\site-packages\sklearn\neural_network\_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.
  warnings.warn(

[LightGBM] [Info] Number of positive: 188, number of negative: 349
[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.000341 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 611
[LightGBM] [Info] Number of data points in the train set: 537, number of used features: 8
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.350093 -> initscore=-0.618630
[LightGBM] [Info] Start training from score -0.618630
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf